A method for estimating worker states using a combination of ambient sensors for remote collaboration

Abstract

Estimation methods that check worker states for remote collaboration have been used in research via wearable sensors, cameras, and microphones, but these methods have some drawbacks. An estimation method that uses both vibration sensors and distance sensors is presented in this research. A prototype module with two sensors is tested to estimate the four states of a user by creating a self-organizing map (SOM) using the sensor data. Tests show that the prototype module estimates the user state by classifying it into one of three clusters, including “key typing” and “leaving a seat,” and others.